embeddings

Definition

An embedding is a fixed-length vector of floats that represents text (or images, or audio) in a high-dimensional space where similar items end up close together. Embedding models from OpenAI, Cohere, Voyage, and open-source projects produce these vectors. Storing them in a vector database (pgvector, Pinecone, Weaviate) enables semantic search and powers RAG pipelines.

When to use

See also

• RAG — Retrieval-Augmented Generation — fetch relevant documents at query time and feed them to the LLM as context.
• LLM — Large Language Model — a transformer-based model trained on internet-scale text to generate and reason.